The Myth of the Absolute-Return Investor M

PERSPECTIVES

The Myth of the Absolute-Return Investor M. Barton Waring and Laurence B. Siegel

n meetings with clients and colleagues in the both sensible and true to the sense of the term past few years, we have noticed that many eluded us. That experience further piqued our otherwise hardheaded and clear-eyed inves- interest in the idea. I tors are excited about “absolute return” So, let us explore the term a bit. It is widely investing. The notion is spreading like wildfire. used, and because words are chosen to a purpose, Many institutional investors have already added, one can find some of that purpose by observing the or are planning to add, an absolute-return “asset context in which a term is used. class” to their policy mix. At a time when pension One important bit of context is that the word funds, foundations, and endowments are under pair “absolute return” has been used most by those pressure to increase their investment returns, abso- managers who resist the practical and theoretical lute-return investing is often positioned as The successes of relative-return investing and who are Answer—with enthusiasts arguing that it will do a looking for terminology that supports their rebel- better job of meeting institutional return require- lious spirit. The term captures this spirit: If ments than other types of investing. benchmark-relative investing is considered inade- The concept seems to have struck a special quate or wimpy by these rebels, then absolute- chord with those who have struggled to fully accept return investing implies that managers can take the the perceived confines of benchmark-relative opposite approach, one that is not benchmark rel- investing. If you have never really understood why ative. (Real men do not use benchmarks!) all investing is, in the end, relative-return investing, What does it mean to be opposite in spirit to then the notion that absolute returns might be supe- relative-return investing? We surveyed websites to rior seems to make good sense. see how investments that are purportedly absolute- Alas, we fear we have disclosed our conclusion return investments are portrayed. Not surpris- in the title, and we have more than a mild suspicion ingly, many of the descriptions are cagey; the web- that our bias shows in this introduction. So what is sites simply use the term without precisely defining absolute-return investing? What is wrong with it; what they mean by it. But some sites are less from where does our skepticism spring? Is there guarded, particularly those outside the United something valuable and redeemable in the concept, States. Here are some samples. and if so, what is it? The first is taken from a well-known financial pundit writing for SmartMoney.com:1 Why Absolute-Return Investing Is But when the bubble burst, and indeed up until this year, simply staying above water has a Myth been perceived as commendable. In fact, The first question is: Just what are “absolute” plenty of managers have boasted of their good returns? We originally assumed we could start this “relative” performance, having lost only single digits, for example, at a time in which the essay by simply reporting an agreed definition of S&P 500 index was down significantly more. the term “absolute-return investing.” Instead, we Of course, I don’t know many groceries that found ourselves repeatedly offering up ideas to can be bought using good “relative” perfor- each other for what the definition would have to be mance, if that performance . . . happens to be for the term to make sense. Definitions that were negative. I don’t have a degree in economics, only a stack of bills to be paid. So I start, perhaps naively so, with the basic notion that a M. Barton Waring is managing director and head of the good year is one in which I make money—end Client Advisory Group at Barclays Global Investors, of story. My benchmark might be low, but it’s San Francisco. Laurence B. Siegel is director of research very strict. A good return is a positive return, in the Investment Division of The Ford Foundation, even at a relatively low level. In hedge-fund New York City. speak, it’s what we call absolute return.

Here is another, from Macquarie Bank, that Definition and Importance of 2 emphasizes higher-than-market, positive returns: Relative Returns Absolute return investments can offer you potential benefits such as: If absolute returns are supposed to be different from—and better than—relative returns, perhaps a •the potential for higher returns than tradi- place to start is with a discussion of relative returns. tional asset classes Beginning in 1963, Sharpe laid the foundation for •the potential to achieve positive returns when how investment professionals understand and traditional share markets are falling— 5 because they often adopt hedging strategies. decompose total returns on portfolios today. His And another, to the same point, from the Aus- work showed how the total return on any portfolio— tralian Stock Exchange:3 note the emphasis—can be decomposed into a part that is ascribable to the return on the market bench- Absolute return funds have the ability to mark, which he called “beta,” and an idiosyncratic— produce positive investment returns regard- in this case, manager-specific—part that is uncorre- less of general market conditions. The strate- lated with the market, which he called “alpha” (plus gies they adopt can produce returns in both 6 rising and falling markets. the risk-free rate, or cash, of course). In the slightly condensed form popularized by These quotations are consistent with the notion Grinold and Kahn (2000a), this relationship is of providing a return pattern that is different in expressed as spirit from those of relative-return investments. If a definition could be teased out of these statements, rp = βprbm + αp. it might be that absolute returns are positive (as in To restate the equation in plain English, the excess absolute value) and always (or at least mostly) bet- return of a manager’s portfolio (excess over the ter than the market. The idea appears to be that, as riskless rate), rp, is the expected beta of that portfo- a result of these combined attributes, the total return lio, βp, multiplied by the excess return of the man- from absolute-return investing will have less down- ager’s normal portfolio or custom benchmark, rbm side risk and more upside return than the total (that is, the risk premium), plus an alpha (or resid- return produced in a relative-return environment. ual term), αp, that is uncorrelated with the beta Many readers of this article will have already return. rejected this expansive approach to defining abso- Sharpe’s observation is perhaps the most pro- lute returns and will have worked to come up with found insight in modern finance. The return on any, their own more sensible and less rebellious defini- repeat any, portfolio consists of a market part and tion. We will offer our own view later—a view that a nonmarket part. In the jargon of finance, this idea will give some comfort to many of those sensible is often abbreviated to simply: A portfolio has a readers. But our view is stronger: We do not sup- part that is beta and a part that is alpha. port continued use of the term. The beta part results from the average future Of course, we would not disagree that a return exposure to total market returns, often expressed in pattern that beats the market in rising markets and terms of one or more market benchmarks. This mix that does not lose money in falling markets would of exposures is sometimes called a “normal portfo- 4 be a good thing—if it really existed ex ante. (And lio.” Most long-only managers know with relative as you will see, something like it can exist—but it clarity what their normal portfolios are; simple, is not an absolute return.) single-factor examples are large-cap value and One more bit of context: Today, the term “abso- fixed-income credit. (Other, more complex exam- lute return” seems to be used most often to describe ples are also possible.) For a purported absolute- what wealthy individual investors have always return manager, the normal portfolio may not have called hedge funds. Perhaps the term is thought to been purposefully or thoughtfully designed—and give more legitimacy or sophistication to the hedge may be more implicit than explicit—but some- fund approach in the institutional context. But where in the manager’s investment style lies a absolute-return investing is really a more general “home,” a set of factor exposures or betas that the term, and it has been applied to alternative strate- manager goes to when he or she has no reason to gies other than hedge funds as well as to certain go somewhere else. conventional long-only managers (particularly Therefore, the notion that every return has a those with concentrated portfolios that bear little beta component and an alpha component applies resemblance to their benchmarks). Here, we focus to any portfolio—that is, to a portfolio with any on hedge funds because that is where the interest normal portfolio or benchmark, including complex is today. multifactor benchmarks. A portfolio that normally

March/April 2006 www.cfapubs.org 15 Financial Analysts Journal contains multiple asset classes (or a fixed-income absolute-return investors: Is the expected return portfolio) could thus be analyzed within the same you offer investors attributable to your expected framework, the only difference being that the betas average exposure to the beta (single or multiple) would represent exposures to style or asset-class that characterizes your normal portfolio, or is it factors other than (only) the equity market.7 attributable to expected alpha generated through Why is this principle important? The returns of skillful beta timing or security selection? (“Both” normal portfolios or custom benchmarks are easily is an admissible answer, but it won’t change our achieved through mixes of index funds or deriva- conclusion.) tive contracts (or, for more exotic market exposures, And we give fair warning: Stop and think care- through some sort of recipe-driven portfolio that is fully before you answer! Here is how the conversa- essentially passive, although perhaps not available tion might go: as an actual index fund8). Thus, beta returns are inexpensive, provide an expected risk premium Is It Beta? If the answer is “beta” (or “both without requiring skill, and are easy to achieve. beta and alpha”), the manager is acknowledg- But positive expected alpha is hard to achieve. ing that returns are attributable in whole (or in A manager must add realized returns over and part) simply to the expected average exposure above the returns of the beta exposures (and above to beta factors—that is, to the fund’s normal the cash return from a zero-beta exposure) to gen- portfolio. So, with this answer, clients could erate pure alpha. The manager’s clients, moreover, get that portion of the return stream very will surely expect pure alpha in subsequent years inexpensively—nearly free relative to hedge and may consider firing the manager if they do not funds or many actively managed products— get it. by holding index funds and various market- The pure alphas result from manager devia- replicating derivatives. The returns are “just tions from the contents of the benchmark through beta,” and as discussed, beta can be purchased security selection or from beta timing.9 (This readily and inexpensively. The “2 and 20” fee principle would apply to a portfolio with a cash for beta is nice work if you can get it! But to benchmark—that is, zero betas—just as it does to a charge clients hedge fund fees for the compo- portfolio with a benchmark consisting of a more nent of returns that could be replicated with an traditional mix of betas.) The original version of index fund is aggressive and probably not sus- Sharpe’s capital asset pricing model presumed that tainable. The fact is, however, that “just beta” the expected alpha would be zero, but for those of may be a more truthful answer than managers us who believe in active management, this alpha are comfortable accepting. Many studies could conditionally have a nonzero positive expected strongly suggest that hedge funds have returns value if the two conditions of inefficiency and skill that are significantly explained by persistent discussed in Waring and Siegel (2003) were satis- beta exposures. fied. Positive realized alphas might well be had sim- So, we’ve warned our absolute-return man- ply through luck, but positive expected alphas ager friends away from giving an answer that require special skill, skill that is sufficient to beat the includes “beta.” They may have us outsmarted, rest of the skillful crowd that plays the markets. however, and plan to answer “alpha.” After all, And because expected alpha depends on a per- everyone knows that the skillful manager with a ception of skill that is agreed between buyers and positive expected alpha has a valuable product that sellers, it is not only hard to achieve but also (quite deserves a substantial fee. So “alpha” sounds like a naturally) expensive. pretty good answer. Right?

Is It Alpha? Be careful. By describing fund Sorting Out Absolute Returns returns as alpha, managers acknowledge that This explanation of betas and alphas is offered for they are relative-return investors! Alpha is a reason: Just because something is called an defined as a relative return, the return generated “absolute-return investment” does not mean it is over and above the manager’s normal beta granted an exception to the first law of financial exposure, or benchmark. gravity described in the previous section: The Thus, we have demonstrated that there is no returns of any portfolio can be broken down into such thing as an absolute-return investor. The market (beta) components and an alpha compo- phrase propagates a myth—a financial air ball, nent. So, here is the money question we are asking cold fusion with other people’s money! Like most all hedge fund managers who fancy themselves myths of active management (see Waring and

Siegel 2005), it is apparently being promulgated to virtue of that skill. Not all hedge fund managers can aid the marketing of yet another cynical investment be above average, so the group as a whole cannot practice—namely, the mixing of alphas and betas at outperform a fair benchmark, but let us celebrate a single fee level (the higher one, naturally). the fact that the best can be expected to outperform Real investing is about understanding the dif- ex ante. ferences between alpha and beta, picking a mix of What we are pointing out is that such manag- betas as the normal portfolio, and trying to add ers depend on skill as any other manager does and alpha to that portfolio through security selection or their goal is not an absolute-return goal but a market-timing bets.10 Real added value comes only relative-return goal, the goal of producing from relative-return investing. expected alpha. If investors understand what is No wonder we could not sensibly define really happening—that all forms of active manage- absolute-return investing: There is no such thing. ment consist of making bets relative to some sort The term is intended to capture investor attention of benchmark—then they have a better chance of by offering an intuitively appealing alternative to identifying managers, including hedge fund man- the disciplines required by relative-return invest- agers, who really do add value. And that under- ing, but at the end of the day, it delivers beta standing must be founded in relative-return space. returns plus or minus relative (alpha) returns. A The universal goal of active management is to add value sensible meaning for the term simply does not over a benchmark. exist, unless one concedes that absolute return Thus, all managers who make the effort to add equals relative return, in which case there is no special value to a portfolio, whether they want to need for the term. It may appear to be a distinct admit it or not, must do the same thing: beat a type of investing, but if there is a distinction, it is benchmark (a normal portfolio or mix of betas). The a distinction without a difference. challenge is the same for a hedge fund, a long-only More specifically, a well-managed hedge fund manager, a market-neutral long–short manager, a is at heart simply a portfolio with a low or zero beta traditional active manager, a quantitative active and a (hopefully) high expected alpha. That’s a manager—whatever type of manager. Even War- relative return, whether the manager wants to ren Buffett has a benchmark, a cost of capital or admit it or not. And, of course, hedge funds are blend of beta payoffs, that he must beat if he wants subject to the same zero-sum-game rules that apply Berkshire Hathaway to go up more than the rest of to all active investing. Carefully designed studies the market.13 So, the most famous absolute-return have found that, as a group and after making rea- investor in the world is, in fact, a relative-return sonable corrections for survival bias, hedge funds investor—as are all absolute-return investors. do not exhibit statistically significant realizations of Relative-return investing may seem timid and 11 alpha. No real surprise here for the hardheaded constrained to those who do not understand the and clear-eyed investor: Hedge funds are not the difference between beta and alpha, but it is the only “magic asset class” that some would like you to means through which real value can be added to believe. Like any other actively managed fund, portfolios. Relative-return investing is the only they rely on special skill for special success. kind of value-added investing that really exists. Get In this light, one can see that hedge funds and over it! other purported absolute-return strategies are not a distinct asset class. As the betas of a hedge fund go to zero, which we will argue is the ideal level of beta The Way Hedge Funds Ought to Be exposure for such a fund, the natural (unconditional We cannot salvage the term “absolute return,” but or market) return is simply the return on cash plus we can salvage the concept of the hedge fund—that or minus the realized alpha. Technically, given that is, of a fund that takes both long and short positions asset classes are always market risk categories, a as originally envisioned by Alfred Winslow zero-beta hedge fund should be in the cash asset Jones.14 To do so requires us to acknowledge that, class, not in some separately named class.12 Saying as we have stated, all efforts to add special value there is an absolute-return asset class is the same as are, at their heart, relative-return investing—a saying there is a pure-alpha asset class, but the search for pure alpha—and success requires meet- industry does not engage in that practice for any ing the two conditions of (1) inefficiency in the other actively managed funds; we put them in the relevant market and (2) skillful selection of invest- asset class that best matches their beta characteristics. ment positions. A hedge fund or any other active Please understand that we are not disparaging manager that is operating in an inefficient market skillful hedge fund managers or any other truly and has special skill at exploiting those inefficien- skillful manager with a positive expected alpha by cies can fairly be expected to add alpha, to beat the

March/April 2006 www.cfapubs.org 17 Financial Analysts Journal great zero-sum game and, therefore, is a thing of folio in many dimensions of beta); its benchmark rareness and value. really is cash—that is, the risk-free rate.17 But hedge funds do have a normal portfolio, a When risk-control technology is used, the set of exposures they go back to when they do not neutrality in beta is, in fact, an expected average have any special insights. Sometimes, hedge funds neutrality across as many market risk factors as are characterized as having a benchmark of cash. possible, up to and including—and, for the best One certainly can imagine a hedge fund for which funds, exceeding—the number of market risk fac- this is appropriate: The normal portfolio for a tors in the models sold by such firms as Barra. The hedge fund with no net expected average expo- term “risk control” is very much evident in the sures to any styles, markets, or other beta factors portfolios built by the most skilled practitioners of could be correctly understood as a zero-beta port- this form of investing. This approach is in striking folio, and its benchmark would be cash.15 contrast to that of traditional hedge fund manag- In fact, when data from actual hedge funds are ers, who, in their resistance to benchmark-relative evaluated, most funds show persistent net positive investing, reject the importance of the difference beta exposures over time. On average, the equity between beta and alpha in their portfolios and see beta of long–short equity hedge funds ranges little or no value in modern risk-control tech- between 0.3 and 0.6, and they also have some beta niques. They completely miss the benefits of this exposure to bonds.16 In effect, most hedge funds technology: If they were using it, the hedge fund normally put fewer dollars into short positions manager (and the fund’s investors) could clearly than into long positions, and their net betas do not distinguish alpha—the result of skill—from beta. completely cancel and go to zero. A high-quality market-neutral long–short fund driven by skillful insights is the highest There is a good reason to have one’s long posi- expression of the art of active management, and it tions offset by short positions in such a manner that represents what hedged investments ought to be.18 they do give a net zero-beta position as the normal But traditional hedge funds have a long way to go portfolio. The reason lies in the proven lack of before they are as desirable an investment as a efficiency of portfolios that are subject to the long- market-neutral long–short fund that is equally only constraint: For a given level of skill, portfolios skillfully managed. As risk-control technologies constrained to be long only deliver only a fraction become more widespread, expect to see the better of the alpha of an unconstrained or market-neutral hedge funds adopt them. portfolio. This principle is explained fully in Grinold and Kahn (2000a, 2000b) and Clarke, de Silva, and Thorley (2002), and it is summarized in Conclusion: Pay Alpha Fees Only Waring and Siegel (2003). for Real Alpha! This observation is not casual: It is one of the We asserted at the beginning that the notion of cornerstones on which modern active management absolute-return investing has seduced many peo- is based. If an investment manager has skill at ple into believing that superior returns can be making investment bets, then that skill is amplified achieved by those with strong views and little or no by incorporating the bets in portfolios that are not regard for benchmarks. But why do people think constrained to hold only long positions or to hold that absolute-return managers exist, and why do any particular amount of beta. The most efficient they think that such (imaginary) managers ought portfolio across a set of buy-and-sell signals is for to earn supercharged returns? the expected average net beta position to be zero Because they want to believe! Beating the market (so that the normal portfolio is zero beta, or cash). is difficult, and in an environment in which respon- The term in the market for this type of strategy sible forecasters envision a 7–8 percent annual has come to be “market-neutral long–short” invest- expected return on equity benchmarks, those who ing. The long–short part of the term captures the want or need a substantially higher return are look- strategy’s hedge fund–like behavior. The marketing for an easy solution, for more return and/or less neutral part makes it clear that a fund following this risk. If they are hiring so-called absolute-return strategy is beta neutral, truly zero beta. It is like a managers or setting up an absolute-return “asset hedge fund in that it has both long and short posi- class,” they must either believe in the magic of the tions, but it is significantly better in that it incorpo- category or be convinced that skill levels are much rates a clear-eyed view of which part of its return higher for hedge fund managers than for the mere is alpha and which part is beta. Think of a hedge mortals who run ordinary long-only funds. But the fund but with modern risk-control technology (so laws of financial gravity have not been abrogated; that it really does have a net-zero-beta normal port- long–short active return is as much a zero-sum

18 www.cfapubs.org ©2006, CFA Institute The Myth of the Absolute-Return Investor game as long-only active return, and a manager this, and active management cannot generate needs special skill—not merely average skill—to returns in this way. A hedge fund will deliver the win the game. It is unlikely that the 8,000-plus risk-free rate plus a beta return related to its normal (mostly newly minted) hedge fund managers are, portfolio plus an alpha return that comes from beta on average, all that much more specially skilled timing, security selection, or whatever. than their long-only counterparts—media hype to So, the term “absolute-return investing” has no the contrary. meaning. It misleads the listener into thinking it has The solution of hiring highly compensated substance that it does not have, and in our opinion, entrepreneurs who do not feel bound by a bench- the term simply should not be used. mark is powerfully marketed. And some of these All investing is about managing a bundle of funds have actually experienced attractive histori- beta and alpha attributes. The investor’s goals are cal returns, which lends support to the faulty con- to understand the beta exposures of the portfolio jecture that absolute-return portfolios are and to pay active fees only when the investor intrinsically a better portfolio design (it is common expects positive alpha—that is, only for benchmark- to confuse realizations with expectations for the beating performance. Managers, including hedge future). What investors actually get when they hire fund managers, with true expected alpha (from one of these would-be absolute-return managers is above-average skill) are hard to find, but they do a variety of market-like or beta exposures (which exist. The investor who wants to invest in a hedge can be hedged away to a net-zero level but which fund needs to keep in mind that a given quantity of rarely are in practice) plus (or minus) positive (neg- skill will have the highest alpha payoff if the man- ative) alphas—as one would obtain with any ager is a market-neutral long–short manager, the investment—minus fees and other costs. And, on modern risk-controlled version of a hedge fund, a average, before fees and costs, the absolute-return version that works hard to have a normal portfolio funds are merely average. that is close to zero net beta. Consider again the notion, from our discussion But whether using the modern incarnations of of defining “absolute return,” that absolute-return the hedge fund or traditional hedge funds, the investing somehow delivers returns that are posi- investor is looking for special skill at beating bench- tive and high regardless of the direction of the mar- marks. By definition, all investors are benchmark- ket. What is wrong with this notion is that it portrays relative investors. absolute-return investing as a magic investment Beating a benchmark is all that matters; it is the approach able to earn outsized total returns with only thing that is worth paying high fees to achieve. little or no risk of negative returns simply because the manager disdains benchmarks and may have a low net This article qualifies for 0.5 PD credit. market exposure (low beta). Markets do not work like

Notes

1. Jonathan Hoenig, “How Are You Doing” (26 November 5. Full disclosure and fair play require us to note that the 2004): www.smartmoney.com/tradecraft/index.cfm?story capital asset pricing model was independently, and =20031126. roughly simultaneously, discovered by several other 2. “Absolute Return Funds,” Macquarie Bank (accessed 6 researchers, but Sharpe has been the most prolific and December 2004): personal.macquarie.com.au/personal/ persuasive exponent of it. investment_strategies/accelerate_wealth/alt_long_term/ 6. In perfectly efficient markets, the expectation for manager absolute_return_funds_detail.htm. alpha is zero. But with some degree of market inefficiency, a manager of above-average skill can have a positive 3. “Absolute Return Funds Fact Sheet,” Australian Stock expected alpha. Realized alpha will always have a substantial Exchange (accessed 7 December 2004): www.asx.com.au/ random component, but for the skillful manager, the mean markets/pdf/HedgeInfoSheet.pdf. of the distribution will have risen. We discuss these issues 4. A return pattern that beats the market in up markets and in Waring and Siegel (2003); we don’t mean to be glib in earns a positive return during down markets could, theo- skipping over some technical details. Our comments apply retically, be achieved with a portfolio consisting of the in the context of the single-index model, the market model, following: the market benchmark plus puts on the market or the capital asset pricing model, with the caveat that we benchmark plus a high expected alpha. (The rub is, of allow for a positive expected value for alpha under the course, the high expected alpha.) Thus, the “strong form” conditions just stated. For an exposition of these closely absolute-return payoff, higher than the market and also related models, see Sharpe, Alexander, and Bailey (1995). positive, could be seen as optionlike. But this payoff is still, 7. Returns-based style analysis—an application of multiple as is obvious from the components required to construct it, regression in which the regressors are the returns on vari- a relative return. ous investment styles or asset-class factors—is the tool most

March/April 2006 www.cfapubs.org 19 Financial Analysts Journal

widely used to determine the historical mix of betas for a skillful. And the academic studies do not support the reality given portfolio, which, in turn, is useful for evaluating the of persistent true alphas from hedge funds. expected future normal portfolio (see Sharpe 1988, 1992). 12. Perhaps not many organizations would put their hedge Holdings-based style analysis may be used to achieve sim- funds in the cash asset class. Because of their highly volatile ilar goals. active returns and their relative lack of liquidity (they are 8. An example of a recipe-driven portfolio, one that provides useless as collateral for a futures position, for example), exposure to exotic beta, might be a merger arbitrage strategy hedge funds simply do not seem to fit this category, even that involves annually buying the top 10 (by market capital- though they do. Asset classes are typically defined as fully ization) acquisition target companies each year and selling diversified market segments (betas), so an investment with the acquiring companies short. Note that such a strategy, a zero beta is technically a cash-asset-class investment. This like most exotic-beta strategies, cannot be completely pas- observation highlights the need for increased proficiency in sive. One cannot know until the end of the year which deals alpha–beta separation by institutional investors. Although will be in the top 10; one needs to figure out when exactly hedge funds most naturally fit in the cash asset class, the to place the trades, and other decisions need to be made. sensible approach is probably to transport the alpha to some Thus, exotic beta cannot typically be delivered at the other asset class. Or hedge funds could simply be consid- extremely low fees that apply to traditional index funds. ered to be in their own category, an asset “class” with a cash- 9. Beta timing is sometimes called “tactical asset allocation” like beta but a highly volatile alpha. (In this case, to be and sometimes (often with an unfairly pejorative tone) consistent, one should separate the alphas in all of the asset “market timing.” Beta-timing decisions consist of moving classes, for full alpha–beta separation.) We in the industry the beta position away from the normal portfolio’s beta confuse ourselves when, as in the case of hedge funds, we position to capture a gain from being underweight when insist on putting a type of investment in the wrong asset the market underperforms or overweight when the market class because of characteristics associated with its alpha. outperforms its equilibrium expectations. 10. Market timing can be regarded as trying to add alpha 13. A cost of capital is simply the expected return on the market relative to a benchmark consisting of a fixed mix of betas (a risks in the portfolio; it is the return to the company’s beta. normal portfolio) by timing among the beta exposures. 14. The investment strategies of hedge fund pioneer Jones were 11. See, in particular, Asness, Krail, and Liew (2001). Using the first described in Loomis (1966) and were cited in Brown ordinary market model, Asness et al. calculated positive, and Goetzmann (2003). Loomis, in private correspondence, significant alphas (t-statistics of +2 or more) for two out of suggested to us that the term “hedge fund” or “hedged nine strategies, negative alphas for two out of nine, and fund” may have originated with Benjamin Graham. positive but not significant alphas for the remaining five. 15. We say “no net expected . . . beta[s]” because hedge fund The aggregate of all nine strategies had an alpha t-statistic managers take both long and short positions and the betas of 0.76. These results are for data from Credit Suisse First from those opposing positions within each beta category Boston/Tremont for the period January 1994 through Sep- may offset each other to a greater or lesser degree. Remem- tember 2000. Using a Dimson–Scholes–Williams adjusted ber that for establishing the normal portfolio, we are focused market model (that is, a model with the led and lagged on the forward-looking expected average beta positions, not market returns as additional regressors), Asness et al. found on deviations from them for market-timing purposes. the alphas to be positive and significant for two of nine 16. See, for example, Asness et al.; Dopfel (2005); Ennis and strategies, positive and insignificant for one, and negative Sebastian (2003); Malkiel and Saha (2005). for the others (Dimson 1979; Scholes and Williams 1977). The aggregate alpha t-statistic was also negative. Market- 17. This portfolio, which starts with zero beta exposures, can neutral equity funds had the highest alpha t-statistics in then be “equitized” or given any beta exposure or mix of both tests. Hedge fund enthusiasts argue that the average beta exposures that the investor wants (using not only hedge fund manager may actually be above average equities) without affecting portfolio efficiency. because the high compensation and freedom of the strategy 18. A manager who makes market-timing (beta-timing) bets attracts the “best” managers. They can then take money could be market neutral and risk controlled. Such a man- away from boring old long-only institutional, retail, and ager’s expected or average betas would be zero over, say, a other “dumb” investors. Not much real meat covers the market cycle—even if at every given moment, the betas of bones of either argument; there are simply too many hedge the portfolio were nonzero. Active beta timing is a legiti- fund managers with too much money under management mate active management discipline, although for some to claim credibly that they are, as a class, extraordinarily technical reasons, it does demand great skill.

References

Asness, Clifford S., Robert Krail, and John Liew. 2001. “Do Dopfel, Fred. 2005. “Waiter, What’s This Hedge Fund Doing in Hedge Funds Hedge?” Journal of Portfolio Management, vol. 28, My Soup? How Hedge Funds Fit into the Institutional Investor’s no. 1 (Fall):6–19. Portfolio.” Investment Insights, Barclays Global Investors, vol. 8, Brown, Stephen J., and William N. Goetzmann. 2003. “Hedge no. 4 (March). Funds with Style.” Journal of Portfolio Management, vol. 29, no. 2 Ennis, Richard M., and Michael D. Sebastian. 2003. “A Critical (Winter):101–112. Look at the Case for Hedge Funds.” Journal of Portfolio Clarke, Roger, Harindra de Silva, and Steven Thorley. 2002. Management, vol. 29, no. 4 (Summer):103–112. “Portfolio Constraints and the Fundamental Law of Active Management.” Financial Analysts Journal, vol. 58, no. 5 Grinold, Richard C., and Ronald N. Kahn. 2000a. Active Portfolio (September/October):48–66. Management. 2nd ed. New York: McGraw-Hill. Dimson, Elroy. 1979. “Risk Measurement When Shares Are ———. 2000b. “The Surprisingly Large Impact of the Long- Subject to Infrequent Trading.” Journal of Financial Economics, Only Constraint.” Investment Insights, Barclays Global Investors, vol. 7, no. 2 (June):197–226. vol. 3, no. 2 (May).

Loomis, Carol. 1966. “The Jones Nobody Keeps Up With.” Sharpe, William F., Gordon J. Alexander, and Jeffery V. Bailey. Fortune (April):237–247. 1995. Investments. 5th ed. Englewood Cliffs, NJ: Prentice-Hall. Malkiel, Burton J., and Atanu Saha. 2005. “Hedge Funds: Risk Waring, M. Barton, and Laurence B. Siegel. 2003. “The and Return.” Financial Analysts Journal, vol. 61, no. 6 Dimensions of Active Management.” Journal of Portfolio (November/December):80–88. Management, vol. 29, no. 3 (Spring):35–51. This article also Scholes, Myron, and Joseph Williams. 1977. “Estimating Betas appeared as part of M. Barton Waring and Laurence B. Siegel from Nonsynchronous Data.” Journal of Financial Economics, “Understanding Active Management,” Investment Insights, vol. 5, no. 3 (February):309–328. Barclays Global Investors, vol. 6, no. 1 (April 2003). Sharpe, William F. 1963. “A Simplified Model for Portfolio ———. 2005. “Debunking Some Myths of Active Management.” Analysis.” Management Science, vol. 9, no. 2 (January):277–293. Journal of Investing, vol. 14, no. 2 (Summer):20–28. This article ———. 1988. “Determining a Fund’s Effective Asset Mix” also appeared as part of M. Barton Waring and Laurence B. (corrected version). Investment Management Review, vol. 2, no. 6 Siegel “Understanding Active Management,” Investment (November/December):59–69. Insights, Barclays Global Investors, vol. 6, no. 1 (April 2003). ———. 1992. “Asset Allocation: Management Style and Performance Measurement.” Journal of Portfolio Management, vol. 18, no. 2 (Winter):7–19.

[ADVERTISEMENT]

March/April 2006 www.cfapubs.org 21